Railway wagon bearing withdrawal automatic production system
By introducing an automated bearing removal production system in the railway truck industry, using AI and robotics to automatically identify, disassemble and classify bearings and bolts, the problems of high labor intensity, inaccurate information recording and high production costs have been solved, and more efficient , more accurate bearing removal process.
Patent Information
- Application Number
- CN202411371811.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-09-29
- Publication Date
- 2024-12-06
AI Technical Summary
The bearing removal lines in the railway truck industry have problems such as high labor intensity, inaccurate information recording, unreasonable process, high risk of damage to bearings and accessories, high production costs and incorrect numbering.
An automated railway truck bearing removal production system was designed, including wheel-to-mount conveyor lines, cleaning stations, visual identification stations, automatic bolt disassembly stations, automatic marking stations and automatic removal stations, using AI technology Identify, disassemble, number and classify bearings and bolts with robots to realize an automated production process.
It improves the identification and recording speed and accuracy of bearings and bolts, reduces manual error rate, reduces the risk of damage to bearings and accessories, optimizes the production process and reduces production costs.
Smart Images

Figure CN119077353A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of railway vehicle maintenance, and in particular to intelligent sorting equipment for bearing unloading machines of railway vehicle axles. Background Art
[0002] At present, the bearing unloading line in the railway freight car industry uses manual wiping of signboards, manual transcription of the date of first installation of the bearing to the outer ring of the bearing, manual recording of signboard information, manual entry of signboard information into the computer, manual wheel conveying, manual use of pneumatic or electric triggers to remove shaft end bolts, manual handling of stacking front covers, manual numbering on the outer ring of the bearing, manual operation of the bearing unloading machine to unload bearings and back blocks, and manual handling of stacked bearings and back blocks. There are problems such as high labor intensity, inaccurate bearing income information recording, unreasonable process, high risk of damage to bearings and accessories, high production cost, easy errors in manual recording of bearing numbers, and complex bearing classification:
[0003] 1. High labor intensity. Currently, three dedicated personnel are assigned to dismantle the stacking front cover, dismantle the stacking shaft end bolts (including anti-loosening sheets, sign boards, and locks), carry the stacking back block, carry the stacking bearings, and write the bearing numbers. Four decomposition inspection personnel enter the bearing income information and wheelset income information, and enter the bearing income information into the HMIS wheel axle station system by checking the actual objects on site, recording the ledger, and entering the information into the computer. The amount of data entered is large and the labor intensity is high.
[0004] 2. There is a waiting phenomenon during the bearing unloading process. At present, the bearing unloading personnel remove the shaft end bolts and then manually push the wheel shaft to the bearing unloading machine to unload the bearing, which causes a serious waiting phenomenon during the process.
[0005] 3. The error rate of inputting bearing income information and wheelset income information is high. Two operators are assigned to input bearing income information. The operators input the bearing income information into the HMIS wheel axle station system by checking the physical objects on site, recording the ledgers, and entering the information into the computer. This results in a large amount of input data and a high error rate of information input.
[0006] 4. The risk of bearings and accessories being damaged is high. The existing bearing removal mainly relies on manual handling and stacking, which can easily cause bearing components to be damaged during the handling and stacking process.
[0007] 5. There are problems of unclear bearing numbers and duplicate bearing numbers. Bearing numbers are numbered by operators using electric engraving pens, which can easily lead to mistakes and unclear numbers during the writing process. Summary of the invention
[0008] The technical problem to be solved by the present invention is: in view of the above-mentioned technical problems existing in the bearing withdrawal line of the current railway freight car industry, an automated production line is provided.
[0009] The technical solution of the present invention:
[0010] An automated production system for unloading and removing bearings of railway freight cars comprises a wheelset conveyor line, which sequentially sends the wheelsets to a cleaning station, a blow-drying station, a visual recognition station, an automatic bolt disassembly station, an automatic marking station and an automatic unloading station; at the cleaning station, high-pressure water is used to clean the shaft end mark plate and the shaft end bolt; the cleaned wheelset is sent to the blow-drying station for drying; at the visual recognition station, the mark plate information at the left and right ends of the axle and the shaft end bolt information are identified, and the bearing repair process and bolt life are automatically determined according to the identified information, and the identification result is sent to the equipment of the shaft end bolt automatic disassembly station; the shaft end bolt automatic disassembly station is provided with a shaft end bolt disassembly device, and a robot stacks the disassembled front cover and bolts; the automatic marking station engraves the bearing maintenance serial number and the first assembly year and month of the bearing on the outer ring of the bearing; after the automatic unloading station locates the wheel axle, a fixture of a corresponding model is automatically matched to automatically unload the bearing, and the robot grabs the bearing and the back stop for classified stacking.
[0011] The wheel pair conveying line is a step-by-step conveying line, and a wheel pushing mechanism and a torsion spring type wheel blocking arm are installed on the side of the rail.
[0012] There are corresponding automation control parameters for different axle models. After the axle is positioned at the cleaning station, the measuring caliper is first raised to detect the axle diameter, and the axle type is determined based on the axle diameter. The axle type is sent to the subsequent automation equipment to select the corresponding control parameters, and then the cleaning equipment 2 is started to clean the mark plate and wheel end bolts.
[0013] After the axle identification station locates the axle, the camera is used to obtain the wheel end image and send it to the AI recognition server. The AI recognition server identifies the information of the mark plates on the left and right ends of the axle and the bolt information on the axle end, and automatically determines the bearing repair period, bolt life and wheelset repair period based on the identified information, and transmits the determination result to the bolt sorting robot 3 at the axle end. The AI recognition server generates the axle serial number of this maintenance and the bearing income serial number, and sends them together with the identified mark plate information to the labeling and labeling integrated machine 6 of the automatic marking station.
[0014] At the automatic bolt disassembly station, the sorting robot 3 first fixes the front cover. After the bolt disassembly device 4 disassembles the bolt, it adsorbs the bolt into the sleeve through suction. The sorting robot 3 adjusts the direction of the fixture to grasp the shaft end bolt, and stacks the bolts and front covers according to waste and different models.
[0015] The automatic unloading station is equipped with a bearing transfer manipulator 8, a bearing pusher 9 and a palletizing robot 10. The bearing pusher 9 automatically requests materials, automatically positions the wheel axle, automatically selects the corresponding model of fixture, and automatically unloads the bearing. The bearing pusher 9, the bearing transfer manipulator 8, the turning mechanism, and the palletizing robot 10 cooperate to complete the grabbing, lifting, transporting, placing, turning, rear block grabbing, and bearing grabbing operations of the bearings on both sides. The palletizing robot 10 automatically completes the classification and stacking operations of the bearing rear block.
[0016] The AI recognition server 11 stores a trained image recognition model, which is trained using a neural network. A pre-classified data set is input into the model for training. The trained image recognition model is used to classify the currently input wheelset photos, and unqualified bolts and end covers are identified. The sorting robot 3 sorts the bolts and end covers at the current workstation to the corresponding stacking location.
[0017] The image recognition model classifies the life of qualified bolts, pre-inputs bolt image data with marked life to train the model, uses the trained image recognition model to classify the currently input bolt image to determine its life, and numbers the bolts and records their life in the system. The sorting robot 3 places the bolts in different stacking locations according to their life.
[0018] Compared with the prior art, the advantages of the present invention are as follows:
[0019] Realize automatic conveying and positioning of axles, realize automatic cleaning of sign plates of different types of axles, realize identification of bolts and sign plate information of different types of axles at the same station, realize disassembly of front covers and bolts of different types of axles at the same station, and stack them by categories, realize marking of different types of bearings at the same station, realize removal of different types of axle bearings and rear blocks at the same station, and stack them by categories.
[0020] Use AI technology to identify signboard information and record it automatically, replacing manual identification and recording. The identification and recording speed is faster, and the rate of download and accuracy is higher.
[0021] AI technology is used to identify the manufacturing year and month of the shaft end bolts, and the bolt life is determined by the manufacturing year and month, replacing manual judgment of the shaft end bolt life, which is more efficient and accurate.
[0022] AI technology is used to automatically identify sign plate information, shaft end bolt marks and axle models, providing a decision-making basis for the automatic sorting and stacking of front covers, shaft end bolts, rear stops and bearings in subsequent processes. BRIEF DESCRIPTION OF THE DRAWINGS
[0023] Figure 1 Production process flow chart.
[0024] Figure 2Structural diagram of the automated production system for unloading bearings from railway freight cars.
[0025] 1- Axle conveyor line 2- Cleaning equipment 3- Bolt sorting robot 4- Bolt disassembly device 5- Sign board visual recognition system
[0026] 6- Labeling and labeling machine 7- Stepping axle conveyor line 8- Bearing transplanting robot 9- Bearing unloading machine 10- Palletizing robot 11- AI recognition server. DETAILED DESCRIPTION
[0027] Example: Figure 1 , the process design of this system is:
[0028] 1. Automatic conveying Automatic conveying axle positioning, bearing unloading automatic production line adopts automatic axle conveyor line which can be used for automatic conveying of RD2 and RE2B axles. The conveyor lines of each process can be operated separately or in conjunction.
[0029] 2. Clean the shaft end mark plate and bolts. The shaft end mark plate and bolt cleaning equipment uses high-pressure water cleaning to clean the surface of the shaft end mark plate and bolts. It can clean the dust, oil and other attachments on the surface of the mark plate and bolts to reveal all the writing. After positioning the axle and before cleaning, a special shaft body detection device is designed to detect the shaft body diameter, determine the shaft type, and pass the shaft type information to the next process.
[0030] 3. Blow dry the shaft end marking plate and bolts, and use high-pressure air to blow dry any remaining water marks on the marking plate and bolts.
[0031] 4. Visually identify the mark plate and bolt information and determine the repair process. The visual system uses a visual camera to take pictures for the identification of the shaft end mark plate and bolt information, and can automatically determine the bearing repair process and bolt life based on the identified information. The image is recognized and classified with the help of a neural network model. A trained image recognition model is stored in the AI recognition server 11. The model is trained using a neural network. A pre-classified data set is input into the model for training. The trained image recognition model is used to classify the currently input wheelset photos and identify the bolts that have expired. The image recognition model classifies the life of qualified bolts. Bolt image data with marked life is input in advance for model training. The trained image recognition model is used to classify the currently input bolt images to determine their life.
[0032] 5. Automatically disassemble and classify and stack bolts, front covers, etc. The automatic disassembler for shaft end bolts and the sorting and stacking robot work together to disassemble and stack the front covers and bolts.
[0033] 6. Automatic laser marking and automatic barcode printing and pasting. The laser automatic marking machine, rim cleaning machine, barcode generation printer and labeling mechanism are combined into an integrated machine for marking the outer ring of the bearing, automatically generating the wheel set serial number and pasting it on the wheel rim.
[0034] 7. Automatically unload and classify the bearings and backstops. The bearing unloading machine, bearing flipping and caching mechanism, seven-axis stacking robot, robot fixture, transfer manipulator, etc. jointly complete the unloading of bearings, and can classify the backstops and bearings into the corresponding material frames.
[0035] 8. Visually identify the shaft end information and verify the information.
[0036] The specific production line layout structure is as follows: Figure 2 The wheelset conveyor line sends the wheelsets to the cleaning station, drying station, visual identification station, automatic bolt disassembly station, automatic marking station and automatic unloading station in sequence. The wheelset conveyor line adopts a torsion spring wheel-blocking arm to ensure that when the conveyor line stops running due to problems, workers can directly push the wheelset to press over the arm and pass directly. When a fault occurs during operation, the arm can also be pressed back under a large external force to prevent damage to the parts of the wheel-pushing arm device. The linear guide is located under the track, and the track can play a primary protection role. A protective ash box is added to the wheel-pushing mechanism to prevent dust and water stains from falling on the linear guide during cleaning. The conveyor line adopts a fully automatic conveying mode, and the wheel axle separated from the bogie enters the step-by-step conveyor line.
[0037] Cleaning station: After the axle is identified, a dedicated measuring caliper is automatically started to detect the diameter of the axle body. After the axle type is determined, the model information is shared with other automated equipment. The measuring caliper includes two calipers that can move laterally, which are installed on a lifting mechanism as a whole. After the axle is detected, it is first raised, and then the two calipers move toward the center until they touch the axle. After the size is obtained, they are retracted and reset. The cleaning program is automatically started to complete the cleaning of the axle end mark plate and bolts. After the high-pressure water cleaning is completed, the axle is automatically pushed into the drying station.
[0038] Drying station: It is automatically completed by a high-pressure air dryer. The equipment automatically issues a request for material, and after identifying the axle, it automatically starts drying. After drying, it automatically pushes the axle to the visual recognition station
[0039] Visual recognition station: It is completed by the wheel axle locator, special industrial camera, AI recognition server 11, etc., and automatically recognizes the information of the sign plates on the left and right ends of the wheel axle and the information of the shaft end bolts. According to the recognized information, it automatically determines the bearing repair process, bolt life (qualified or unqualified), and wheelset repair process, and transmits the judgment results to the shaft end bolt sorting robot to guide the robot to place the bolts. The wheel axle serial number and bearing collection serial number are automatically generated according to the rules, and the recognized sign plate information is classified and transmitted to the laser automatic marking and barcode automatic printing and pasting integrated machine.
[0040] Automatic bolt disassembly station: The process is completed by the sorting robot 3 and the bolt disassembly device 4, in which the robot fixture can grab and place the bolts and the front cover at the same time without replacement. The sorting robot 3 first fixes the front cover, and the bolt disassembly device 4 removes the bolt and absorbs the bolt into the sleeve through suction. The sorting robot 3 adjusts the fixture direction to grab the shaft end bolt, and the robot automatically completes the classification of bolts and front covers (waste, M22 qualified bolts, RD2 front covers, M24 qualified bolts, RE2B front covers) and stacking.
[0041] Automatic marking station: It is automatically completed by the marking and labeling integrated machine 6 and the axle positioning device. The marking and labeling integrated machine 6 is a laser automatic marking, barcode automatic printing and pasting integrated machine. The equipment automatically engraves the QR code, the bearing collection serial number and the repair process mark on the outer ring of the bearing, and stores the bearing inspection serial number and the year and month of the first installation of the bearing in the QR code. The equipment automatically prints the axle barcode, automatically cleans the rim, and automatically pastes the axle barcode. The equipment automatically transports the axle to the bearing unloading machine and automatically completes the material request operation.
[0042] Automatic unloading station: bearing unloader 9, bearing transfer manipulator 8, flip mechanism, and stacking robot 10 jointly complete the process. The bearing unloader 9 automatically requests materials, automatically positions the axle, automatically selects the corresponding model of fixture, and automatically unloads the bearing. The bearing unloader 9, bearing transfer manipulator 8, flip mechanism, and stacking robot 10 cooperate to complete the grabbing, lifting, transporting, placing, flipping, rear block grabbing, and bearing grabbing operations of the bearings on both sides. The stacking robot 10 automatically completes the classification and stacking of the bearing rear block.
Claims
1. An automated production system for unloading bearings of railway freight cars, comprising an axle conveyor line, characterized in that: The axle conveyor line sends the axle to the cleaning station, drying station, visual recognition station, automatic bolt disassembly station, automatic marking station and automatic unloading station in sequence; the mark plate and shaft end bolts are cleaned with high-pressure water at the cleaning station; the cleaned axle is sent to the drying station for drying; the information of the mark plates on the left and right ends of the axle and the shaft end bolts are identified at the visual recognition station, and the bearing repair process and shaft end bolt life are automatically determined based on the identified information, and the identification results are sent to the equipment of the shaft end bolt automatic disassembly station; the shaft end bolt automatic disassembly station is equipped with a shaft end bolt disassembly device, and the disassembled front cover and shaft end bolts are stacked by the robot; the automatic marking station engraves the bearing maintenance serial number and the year and month when the bearing was first installed on the outer ring of the bearing; the automatic unloading station locates the axle and automatically matches the corresponding model of fixture to automatically unload the bearing. After the bearing is unloaded, the transfer robot is used to transport the bearing to the rear block disassembly station, and the robot then grabs the bearing and rear block for classification and stacking.
2. According to claim 1, the railway freight car bearing unloading automated production system is characterized by: The wheel pair conveyor line is a step-by-step conveyor line, and a wheel pushing mechanism and a torsion spring type wheel blocking arm are installed on the side of the rail.
3. The railway freight car bearing unloading automated production system according to claim 2 is characterized by: There are corresponding automation control parameters for different axle models. After the axle is positioned at the cleaning station, the measuring caliper is raised first to detect the axle diameter. The axle type is determined based on the axle diameter, and the axle type is sent to the subsequent automation equipment. The corresponding control parameters are selected, and then the cleaning equipment is started to clean the mark plate and the axle end bolts.
4. The railway freight car bearing unloading automated production system according to claim 3 is characterized by: After the axle identification station locates the axle, the camera is used to obtain the images of the mark plate and the head of the shaft end bolt and send them to the AI identification server (11). The AI identification server (11) identifies the mark plate information at the left and right ends of the axle and the shaft end bolt information, automatically determines the bearing repair process and bolt life based on the identified information, and transmits the determination result to the shaft end bolt sorting robot (3). The AI identification server (11) generates the axle serial number and the bearing collection serial number of this maintenance, and sends them together with the identified mark plate information to the labeling and labeling integrated machine (6) of the automatic marking station.
5. The railway freight car bearing unloading automated production system according to claim 4 is characterized by: At the automatic disassembly station for the shaft end bolts, the sorting robot (3) first fixes the front cover, and the bolt disassembly device (4) removes the bolts and then absorbs the bolts into the sleeve by suction. The sorting robot (3) adjusts the direction of the fixture to grasp the shaft end bolts, and the shaft end bolts are automatically sorted according to waste and bolts to be repaired, and the front covers are sorted and stacked according to different models.
6. The railway freight car bearing unloading automated production system according to claim 5, characterized in that: The automatic bearing unloading station is provided with a bearing transfer manipulator (8), a bearing unloading machine (9) and a palletizing robot (10). The bearing unloading machine (9) automatically loads materials, automatically positions the wheel axle, automatically selects a fixture of a corresponding model, and automatically unloads the bearing. The bearing unloading machine (9), the bearing transfer manipulator (8), the turning mechanism, and the palletizing robot (10) cooperate to complete the grabbing, lifting, transporting, turning, and placing operations of the double-sided bearing backstops. The palletizing robot (10) automatically completes the classification and stacking operations of the bearings and backstops.
7. The railway freight car bearing unloading automated production system according to claim 6, characterized in that: The AI recognition server (11) stores a trained image recognition model, which is trained using a neural network. A pre-classified data set is input into the model for training. The trained image recognition model is used to recognize the currently input sign board photo, and the information engraved on the sign board can be automatically recognized. The repair process of the bearing is determined based on the information on the sign board. The sorting robot (3) sorts the front cover of the current workstation to the corresponding stacking location.
8. The railway freight car bearing unloading automated production system according to claim 7, characterized in that: The manufacturing year and month of the shaft end bolts are identified by an image recognition model. The bolt head mark image data is input in advance to train the model. The trained image recognition model is used to identify the manufacturing year and month of the shaft end bolts for the currently input bolt image. The life of the shaft end bolts is determined based on the identified manufacturing year and month of the shaft end bolts. The sorting robot (3) places the bolts in different stacking locations according to their life.
Citation Information
Patent Citations
Bearing assembly cleaning system
CN114135587A
Intelligent disassembling and sorting system for bearing end covers
CN117961505A
Rail vehicle intelligence mechanical arm withdrawal transmission line
CN208467705U
Automatic withdrawing device for railway wagon bearing
CN223394793U
Product automatic registration in distribution center using artificiality vision recognition based on cloud and server performing thereof
KR102493942B1
Cited By
Product mixed flow disassembling line with dynamic classification and secondary reflow and control method
CN122789128A
